Engineering bible
The Agent OS engineering corpus.
169 versioned specifications with stable identifiers, explicit status and evidence requirements across product vision, architecture, hardware, security, research, legal controls and execution planning.
Specifications
Explore by discipline.
Product vision, architecture, hardware, research, legal controls, governance, decisions and execution plans form one cross-linked body of work.
169 documents
AGENT-OS-WIDER-LENSAgent Os Wider LensResearchAGENTOS-SPEC-DIGEST-PRODUCT-ARCHITECTUREAgentos Spec Digest Product ArchitectureResearchAOS-ADR-0001ADR-0001: Fork Fuchsia/Zircon as the Microkernel BaseADR-0001: Fork Fuchsia/Zircon as the microkernel base: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme. Zircon, DFv2, FIDL, Magma, and Starnix are taken from Fuchsia; board drivers and the Rust-first product layer are built on top.DecisionsAOS-ADR-0002ADR-0002: Portable Layer ContractsADR-0002: Portable Layer Contracts: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.DecisionsAOS-ADR-0003ADR-0003: Parallel Hardware TracksADR-0003: Parallel Hardware Tracks: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.DecisionsAOS-ADR-0004ADR-0004: Minimal Android/Linux Use on Pixel 9ADR-0004: Minimal Android/Linux Use on Pixel 9: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.DecisionsAOS-ADR-0005ADR-0005: Data and Backup ClassesADR-0005: Data and Backup Classes: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.DecisionsAOS-ADR-0006ADR-0006: Working Name and Trademark ClearanceADR-0006: Working Name and Trademark Clearance: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.DecisionsAOS-ADR-0007ADR-0007: Archive the Pixel 9 Native RouteADR-0007: Archive the Pixel 9 Native Route: the Pixel 9 native-feasibility track is moved to archived status with documentation retained, quarantine rules unchanged, and explicit reactivation conditions.DecisionsAOS-ARCH-001Portable System ArchitectureThe normative layered architecture, portability boundaries, dependency rules, degradation model, and future-device readiness.ArchitectureAOS-ARCH-002Microkernel SpecificationScope, object model, syscall discipline, unsafe policy, fault behavior, determinism, and acceptance for the forked Fuchsia/Zircon microkernel.ArchitectureAOS-ARCH-003Process, Memory, and Scheduling ModelProcess, Memory, and Scheduling Model: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-004Capabilities and IPCCapabilities and IPC: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-005IDL, API, and VersioningIDL, API, and Versioning: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-006Drivers, HAL, and Board PackagesIsolated driver domains, board packages, hardware-service contracts, DMA/interrupt security, firmware provenance, and lifecycle behavior.ArchitectureAOS-ARCH-007System ServicesSystem services, including first-class routing, tethering, VPN gateway, policy, failure, evidence, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-008Graphics, Compositor, and UI RuntimeGraphics, Compositor, and UI Runtime: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-009Storage, Entity Graph, History, Sync, and Bundle StateDurable entities, semantic history, CRDT and transactional state, micro-app instances, encrypted sync, and delay-tolerant Agent Mesh bundle state.ArchitectureAOS-ARCH-010Agent Runtime, Micro-App Generation, and Action SafetyAgent planning, capability grants, text-to-micro-app generation, provider and route selection, effect receipts, budgets, confirmation and bounded autonomy.ArchitectureAOS-ARCH-011Compatibility and Pixel 9 Legacy PolicyStrict quarantine of Android/Linux to Pixel 9 research, temporary sidecar rules, trace replay, retirement records, and future compatibility admission.ArchitectureAOS-ARCH-012Security Architecture and Threat ModelSecurity Architecture and Threat Model: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-013Boot, Update, and RecoveryBoot, Update, and Recovery: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-014Power, Thermal, and Background ExecutionPower, thermal, non-blocking background execution, durable transfer, shutdown handoff, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-015Build, Test, Evidence, and ObservabilityBuild, Test, Evidence, and Observability: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-016Layered Portability and Backend ConformanceLayered Portability and Backend Conformance: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-017Kernel Boot and Bring-up SequenceKernel Boot and Bring-up Sequence: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-018Legacy Containment and ReplacementLegacy Containment and Replacement: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-019Formal Models and Assurance RoadmapFormal Models and Assurance Roadmap: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-020Portable Device-Service ContractsPortable Device-Service Contracts: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ArchitectureAOS-ARCH-021Layered Boot and Instant ModesFounder-directed architecture: the OS boots in layers, and the device can (re)boot near-instantly into primitive single-purpose modes — typewriter, reader, calculator, remote — as both a battery strategy and a focus feature.ArchitectureAOS-ARCH-022Layer Manifest Format and Capability ModelThe concrete format that makes ARCH-021 layered boot and instant modes buildable: a declarative layer/mode manifest and the capability model that enforces radio-less modes and scoped agent authority by construction.ArchitectureAOS-ARCH-023Build Strategy and Developer Experience for the Fuchsia ForkHow we neutralize the heavy Fuchsia full-tree build: a two-tier model where day-to-day product work uses the Bazel SDK against prebuilt platform artifacts (no platform build at all), and only kernel/driver deltas require the full GN/ninja tree; incremental ninja + ccache make the full tree a first-build-only cost; ApplArchitectureAOS-ARCH-024Agent Mesh: Multi-Transport Peer ConnectivityTransport-neutral peer-to-peer connectivity across local IP, Wi-Fi Direct, BLE, LoRa, USB, gateways, and intermittent store-and-forward paths.ArchitectureAOS-ARCH-026Micro-App Runtime and Declarative Surface ContractsRuntime, manifest, component, action, state, sandbox, lifecycle, and multi-surface conformance model for Agent OS micro-apps.ArchitectureAOS-BRIEFExecutive BriefingDecision-oriented summary of the product thesis, Fuchsia/Zircon-fork kernel, portable layers, hardware routes, legal controls, first actions, purchases, budget, and evidence gates.GeneralAOS-GLOSSARYGlossary and Alphabetical IndexDetailed definitions, significance, examples, and exact specification references for 229 specialized Agent OS terms.GlossaryAOS-GOV-001Specification Lifecycle and Change ControlSpecification Lifecycle and Change Control: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.GovernanceAOS-GOV-002Contribution and Community ModelContribution and Community Model: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.GovernanceAOS-GOV-003Funding, Grants, and BountiesFunding, Grants, and Bounties: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.GovernanceAOS-GOV-004Security Disclosure and ResponseSecurity Disclosure and Response: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.GovernanceAOS-GOV-005Community Launch and Maintainer FormationCommunity Launch and Maintainer Formation: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.GovernanceAOS-GOV-006Contact Acquisition and Partnership PlaybookContact Acquisition and Partnership Playbook: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.GovernanceAOS-HW-001Hardware Target PortfolioParallel hardware target strategy, admission criteria, scoring dimensions, minimum lab portfolio, and retirement rules.HardwareAOS-HW-002Pixel 9 Quality-Hardware TrackPixel 9 Quality-Hardware Track: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-003Open Phone-Form-Factor TrackOpen Phone-Form-Factor Track: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-004Documented Development-Board TrackDocumented Development-Board Track: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-005Semi-Open Quality Hardware TrackSemi-Open Quality Hardware Track: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-006Camera Quality ProgramMulti-path program for achieving good camera quality without coupling product architecture to a proprietary smartphone ISP.HardwareAOS-HW-007Cellular and Telephony ProgramCellular and Telephony Program: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-008Custom Device and ODM ReadinessArchitecture, product identity, evidence, commercial package, contracts, and prototype stages required before a future ODM or contract-manufactured device.HardwareAOS-HW-009Lab, Procurement, and Initial BudgetInitial laboratory, hardware, legal, and specialist procurement waves with budget ranges, safety controls, and inventory requirements.HardwareAOS-HW-010Hardware Candidate ScorecardHardware Candidate Scorecard: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-011Open, Documented, and Reversible Hardware CatalogOpen, Documented, and Reversible Hardware Catalog: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-012Camera Hardware Options and Quality PathCamera Hardware Options and Quality Path: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-013Hardware Compromise Elimination StrategyHardware Compromise Elimination Strategy: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-014Pixel 9 Evidence and Reverse-Engineering DossierPixel 9 Evidence and Reverse-Engineering Dossier: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-015Debug, Power, Camera, and Reliability LaboratoryDebug, Power, Camera, and Reliability Laboratory: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-016Vendor Documentation and Support AcquisitionVendor Documentation and Support Acquisition: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.HardwareAOS-HW-017Interim Demo Device: Module BrickA pocketable demonstration handheld assembled from documented, pre-certified modules: reference architecture, baseline BOM, cellular and camera strategy, demo-claim limits, and intake tasks.HardwareAOS-HW-018Demo Brick V1 Configuration BaselineFrozen V1 configuration for the interim demo brick: CM5 compute with native (non-Linux) OS independence notes, portless industrial design, soldered consumer eUICC (true eSIM), NVMe storage expansion, SurfaceVolume gesture control, Taptic-class haptics-as-audio with software prior art, IR blaster, optional sensor bay, fHardwareAOS-HW-018-BOMDemo Brick V1 — Bill of Materials (Appendix A to AOS-HW-018)Store-grouped shopping list for one Demo Brick V1 unit: five stores cover everything. Observed prices and direct links (2026-07); frozen SKUs recorded by AOS-DEMO-001.HardwareAOS-HW-019Power Architecture and Ultra-Low-Power StandbyHeterogeneous power architecture for the demo brick: an always-on microcontroller domain (always-on island) that hard-gates the application processor and radios, enabling days-long standby and a competitive energy story rather than the hours-long idle of a bare CM5.HardwareAOS-HW-020Device Catalogue and Per-Target MatrixPer-target device catalogue: role, quality ceiling, documentation access, boot control, GPU, camera, modem, power, security, debug, firmware, legal route, cost, availability, replacement, and go/no-go for every candidate the programme tracks, aligned with ADR-0007 (Pixel 9 hardware archived, demo brick current) and theHardwareAOS-HW-021LoRa and Low-Power Mesh Hardware TrackExternal, accessory, fixed-relay, and future integrated radio options for Agent Mesh.HardwareAOS-INDEXDocumentation IndexWiki-style map of all canonical documents, ten volumes, source digests, datasets, and recommended reading paths.GeneralAOS-LEGAL-001Legal and Compliance FrameworkLegal and Compliance Framework: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-002Clean-Room Reverse Engineering ProtocolClean-Room Reverse Engineering Protocol: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-003Open Source, IP, and Provenance PolicyOpen Source, IP, and Provenance Policy: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-004Regulatory and Certification RoadmapRegulatory and Certification Roadmap: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-005Contact and Partnership MapPrioritized counsel, silicon, phone, camera, modem, lab, assurance, funding, and ODM contact plan with precise asks.LegalAOS-LEGAL-006Trademark, Brand, and Public CommunicationsTrademark, Brand, and Public Communications: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-007Vendor NDAs, Evaluation Agreements, and Device ContractsVendor NDAs, Evaluation Agreements, and Device Contracts: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-008Reverse-Engineering Jurisdiction MatrixReverse-Engineering Jurisdiction Matrix: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-009Community Contributions, Licensing, and IP GovernanceCommunity Contributions, Licensing, and IP Governance: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-010Data Protection, Agent Processing, and Privacy EngineeringData Protection, Agent Processing, and Privacy Engineering: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-011Certification, Carrier, Security, and Compliance Contact PlanCertification, Carrier, Security, and Compliance Contact Plan: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.LegalAOS-LEGAL-012Prototype Travel and Border ConsiderationsPractical and regulatory considerations for carrying the demo brick across borders: security screening, radio and encryption regimes, IMEI registration, data-at-border hygiene, and the travel kit protocol. Not legal advice.LegalAOS-LEGAL-013Consolidated Threat Model and Security PostureOne place that states what the demo brick and AgentOS defend against, what they explicitly do not, the honest limits of the CM5 platform (closed boot ROM, no verified boot), and the security properties that hold anyway — so 'privacy-first' is a claim with evidence, not a slogan.LegalAOS-LEGAL-014Per-Component License StrategyComponent-by-component licensing decision for Agent OS, shaped by the fact that we fork the entire Fuchsia tree (Zircon MIT-style; userspace BSD/Apache-2.0 with a patent grant). Because Fuchsia is fully permissive, our own code may be licensed freely; this document chooses per component by business and community model LegalAOS-LEGAL-015Regional Radio and Mesh ComplianceRegion, certification, transmit, duty-cycle, emergency, privacy, and product-claim controls for Agent Mesh radios.LegalAOS-MANIFESTRelease ManifestInventory, validation state, reproducibility rules, canonical datasets, and release boundaries for the Agent OS Engineering Bible.GeneralAOS-META-001Linking and Publishing ConventionLinking and Publishing Convention: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.metaAOS-META-002Document Schema and MetadataDocument Schema and Metadata: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.metaAOS-META-003Source and Claim Citation ConventionSource and Claim Citation Convention: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.metaAOS-META-004Repository LayoutRepository Layout: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.metaAOS-PLAN-001Program Model and Parallel TracksProgram Model and Parallel Tracks: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-002Roadmap and Gantt BaselineA multi-year, parallel, evidence-gated roadmap with Mermaid Gantt and milestone exit evidence.PlanningAOS-PLAN-003Work Breakdown StructureWork Breakdown Structure: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-004Budget and Resourcing ModelBudget and Resourcing Model: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-005Risk RegisterRisk Register: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-006Decision Gates and Stop CriteriaDecision Gates and Stop Criteria: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-007First 90 DaysConcrete first-week, first-month, 30–60 day, and 60–90 day execution sequence and exit evidence.PlanningAOS-PLAN-008GitHub Operating ModelGitHub as the single public source of truth (Issues + Project v2) with a reproducible import/reconciliation process. Linear is not used.PlanningAOS-PLAN-009Task CSV Data Dictionary and Agent InstructionsCanonical task CSV schema, writing standard, import views, cross-reference format, agent sequence, and validation rules.PlanningAOS-PLAN-010Nine-Phase Execution ModelNine-Phase Execution Model: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-011Ten-Volume Documentation ProgrammeTen-Volume Documentation Programme: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-012Cross-Track Dependency and Convergence MapCross-Track Dependency and Convergence Map: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-013First Purchases, Spend Gates, and 90-Day BudgetFirst Purchases, Spend Gates, and 90-Day Budget: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-014Linear and GitHub Import and Reconciliation RunbookLinear and GitHub Import and Reconciliation Runbook: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-015Review, Briefing, and Decision CadenceReview, Briefing, and Decision Cadence: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-016Agent Execution Brief for Documentation and Tracker AutomationAgent Execution Brief for Documentation and Tracker Automation: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.PlanningAOS-PLAN-017Demo Brick Task Intake (AOS-DEMO family)Formal intake of the AOS-DEMO task family into the catalog: IDs, priorities, dependencies, and verification for the interim demo brick programme (HW-017/018/019, ARCH-021).PlanningAOS-PLAN-018Demo Brick Kickoff Sequence (first 30 days)Four parallel lanes for the first month: procurement (longest lead times first), no-hardware software, evidence bench, and documentation reconciliation.PlanningAOS-PLAN-019Agent Mesh and Micro-App Execution PlanTask, dependency, evidence, and gate plan for Agent Mesh and the text-to-micro-app platform.PlanningAOS-PROD-001Entity-First Shell and SurfacesEntity-first shell, project context, named workspaces, surfaces, actions, history, consent, recovery, and provider contracts.ProductAOS-PROD-002Global History and Semantic JournalingGlobal History and Semantic Journaling: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-003Actions, Integrations, Widgets, and Micro-AppsTyped actions, provider integrations, declarative surfaces, and cross-surface micro-apps for the APP-LAST product model.ProductAOS-PROD-004Backup, Migration, and User RecoveryBackup, Migration, and User Recovery: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-005Accessibility, Localization, and InputAccessibility, Localization, and Input: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-006Stock Experiences and Product Surface RoadmapStock Experiences and Product Surface Roadmap: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-007IntentBox and FlexLightIntentBox and FlexLight: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-008Malleable and Itemized Software ModelMalleable and Itemized Software Model: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-009Personal Data Authority and Agent AccessPersonal Data Authority and Agent Access: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-010Transclusion, SideMemo, and Contextual LinkingTransclusion, SideMemo, and Contextual Linking: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-011Action and Provider InteroperabilityAction and Provider Interoperability: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-012Product Surface Tiers and Dependency RulesProduct Surface Tiers and Dependency Rules: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ProductAOS-PROD-013Compute Subscription and Elastic OffloadFounder-directed monetization and system feature: subscription tiers sell not only cloud storage but server compute; the OS opportunistically offloads heavy work to subscribed server resources over stable links, saving battery, thermals, and hardware cost — without breaking the local-first contract.ProductAOS-PROD-014Native App Clients and Service IntegrationHow popular apps and banking services reach Agent OS: a compatibility layer as the floor and first-class native ports as the goal, with a strict legality ladder — official/public APIs and licensed programs first, clean-room reverse engineering where lawful, and a hard line against anything unlawful (no anti-tamper/pinnProductAOS-PROD-015Voice Agent and Conversational InterfaceThe voice agent is the primary differentiator that justifies the far-field audio hardware: an inspectable, capability-scoped, local-first conversational interface to the entity/action/history system, with a tiered on-device/server split and honest confirmation of every effectful action.ProductAOS-PROD-017Off-Grid Peer-to-Peer ExperienceUser-facing Agent Mesh behavior for direct, delayed, relayed, emergency, and offline delivery.ProductAOS-PROD-018Text-to-Micro-App and Widget BuilderA text-first, Shortcuts-like builder for safe interactive micro-apps rendered across documents, AI responses, notifications, home/lock surfaces, wearables, vehicles, and focused modes.ProductAOS-READMEAgent OS Engineering BibleCross-linked architecture, hardware, research, legal, planning, glossary, evidence, and import package for a Fuchsia/Zircon-fork portable operating system.GeneralAOS-RES-001Prior Art and Literature AtlasAnnotated atlas of operating systems, capability security, Rust OS work, mobile hardware, camera, cellular, and supply-chain prior art.ResearchAOS-RES-002Hardware Evidence and Source RegisterPrimary and project sources used to support architecture, hardware, legal, product, and programme decisions.ResearchAOS-RES-003Claim Verification RegisterClaims are classified as decisions, supported findings, hypotheses, bounded inferences, contradictions, or unresolved questions.ResearchAOS-RES-004Experiment and Evidence CatalogExperiments turn material uncertainty into bounded, reproducible evidence and explicit decision gates.ResearchAOS-RES-005Source Corpus, Authority, and Conflict MapSource Corpus, Authority, and Conflict Map: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ResearchAOS-RES-006Agentic Platforms, Local-First Systems, and Malleable SoftwareAgentic Platforms, Local-First Systems, and Malleable Software: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ResearchAOS-RES-007Microkernel and Portability Prior-Art SynthesisMicrokernel and Portability Prior-Art Synthesis: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ResearchAOS-RES-008GPU, Camera, Cellular, Power, and Boot Prior ArtGPU, Camera, Cellular, Power, and Boot Prior Art: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ResearchAOS-RES-009Literature and Prior-Art Reading ProgrammeLiterature and Prior-Art Reading Programme: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ResearchAOS-RES-010Claim-to-Experiment MatrixClaim-to-Experiment Matrix: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ResearchAOS-RES-011Interim Hardware Market Survey and Demo BOM SourcingSurvey of DIY/open phone prior art (ZeroPhone, WiPhone, CM4/CM5 handhelds, OnePlus donors, postmarketOS, FLX1), the recommended demo-brick configuration, and a sourced bill of materials with Moscow and Serbia/EU purchase channels and observed prices.ResearchAOS-RES-012Fuchsia/Zircon-Fork Specification — Engineering DigestDigest of the 60-page specification for the chosen approach: forking Fuchsia/Zircon (taking Zircon, DFv2, FIDL, Magma, Starnix as-is), writing the Rust-first entity/agent product layer from scratch, and bringing hardware up on a separate track. The Pixel-9 hardware target within it is archived (ADR-0007) in favor of thResearchAOS-RES-013OS Comparison Matrix — iOS vs Android vs Agent OSSourced, dimension-by-dimension comparison of iOS, Android, and Agent OS across system object model, action/intent APIs, agent authority, global history, sandboxing/capabilities, update model, privacy, offline behaviour, extensibility, interop, identity, and developer distribution. Every iOS/Android cell is backed by aResearchAOS-RES-014System Survey: What Agent OS Takes From Twenty-One Prior Operating Systems and Data SubstratesA sourced, one-section-each survey of the operating systems, verified kernels, embedded runtimes, action frameworks, and CRDT substrates that most shape Agent OS. Each system states what it is, the one to three ideas that matter to us, and an explicit takeaway for the Fuchsia-fork product line: what to adopt, what to sResearchAOS-RES-015AgentOS Shell UI Stack — Options, Feasibility, RecommendationA decision-oriented, sourced report on how to build the AgentOS shell and UI stack on a Fuchsia-fork base. It surveys the native Fuchsia UI options (Flatland, Carnelian, abandoned Flutter-on-Fuchsia, Chromium/WebEngine, Starnix), evaluates a Swift/SwiftUI-style path and a React/flexbox path, assesses the Flutter fork (ResearchAOS-RES-016AgentOS Design System & Component StrategyA decision-oriented, sourced report on the design-system, component-library, and Rust-GUI-framework decision for the AgentOS shell, synthesising three independent research rounds. It separates two independent decisions: the visual design LANGUAGE (DECIDED by the owner: shadcn/ui, MIT — Radix-primitive semantics plus a ResearchAOS-RES-017Mesh, DTN, and Off-Grid Communication Prior ArtPrior-art map for Meshtastic, Reticulum, MeshCore, Briar, DTN Bundle Protocol, Wi-Fi Aware, BLE mesh, LoRaWAN, and opportunistic networking.ResearchAOS-RES-IOS-VS-ANDROID-SUMMARY-ENiOS vs Android vs Agent OS — English SummaryEnglish distillation of the Russian grievance register comparing iOS and Android: where Android still trails iOS (with 2026 updates that closed some gaps), where Android is quietly better, and the core thesis that both are exhausted at the app-centric ceiling — the product motivation for Agent OS.ResearchAOS-RES-PRIOR-ART-ATLAS-SUMMARY-ENPrior Art Atlas — English SummaryEnglish distillation of the Russian prior-art atlas: for each Agent OS concept (entity-first data, cards/widgets over apps, agents, global history), the historical precedents that tried it, why they died, and the surviving lesson — condensed into six failure rules and a modern (2025–2026) agentic-front scan.ResearchAOS-RES-WIDER-LENS-SUMMARY-ENAgent OS Wider Lens — English SummaryEnglish distillation of the Russian wider-lens survey: who is already digging the same tunnels (platform agent layers, local-first components, malleable software, itemized OS, global history products), which components are ready to reuse, the gaps our notes miss, and the concrete spec edits that follow.ResearchAOS-TASKSCanonical Task CatalogHuman-readable projection of the canonical task CSV, including descriptions, acceptance criteria, dependencies, evidence, and traceability.PlanningAOS-TPL-ADRArchitecture Decision Record TemplateArchitecture Decision Record Template: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.templatesAOS-TPL-CONTACTContact Outreach TemplateContact Outreach Template: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.templatesAOS-TPL-EVIDENCEHardware Evidence Record TemplateHardware Evidence Record Template: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.templatesAOS-TPL-EXPERIMENTExperiment Record TemplateExperiment Record Template: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.templatesAOS-TPL-SPECSpecification TemplateSpecification Template: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.templatesAOS-VAL-001System and Backend Conformance StrategySystem and Backend Conformance Strategy: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ValidationAOS-VAL-002Evidence Bundle and Experiment SchemaEvidence Bundle and Experiment Schema: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ValidationAOS-VAL-003Documentation, Link, Source, and Task ValidationDocumentation, Link, Source, and Task Validation: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ValidationAOS-VAL-004Hardware Test and Quality MatrixHardware Test and Quality Matrix: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ValidationAOS-VAL-005Security, Privacy, and Agent Assurance PlanSecurity, Privacy, and Agent Assurance Plan: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.ValidationAOS-VALIDATIONRelease Validation ReportValidation scope, counts, integrity results, and known limitations for the packaged foundation release.ValidationAOS-VOL-01Volume 01: FoundationVolume 01: Foundation: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-02Volume 02: Kernel and Core SystemVolume 02: Kernel and Core System: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-03Volume 03: Hardware and Device ProgramVolume 03: Hardware and Device Program: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-04Volume 04: Research and Prior ArtVolume 04: Research and Prior Art: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-05Volume 05: Legal and ComplianceVolume 05: Legal and Compliance: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-06Volume 06: Programme and Project ManagementVolume 06: Programme and Project Management: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-07Volume 07: Procurement and LaboratoryVolume 07: Procurement and Laboratory: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-08Volume 08: Contacts, Partnerships, and CommunityVolume 08: Contacts, Partnerships, and Community: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-09Volume 09: Knowledge Base and Product VocabularyVolume 09: Knowledge Base and Product Vocabulary: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VOL-10Volume 10: Validation, Assurance, and PublicationVolume 10: Validation, Assurance, and Publication: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.volumesAOS-VSN-001Product VisionAPP-LAST product thesis: durable entities, shared actions, semantic history, generated micro-apps, local authority, Agent Mesh peer connectivity, and inspectable agents.VisionAOS-VSN-002Principles and Non-GoalsPrinciples and Non-Goals: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.VisionAOS-VSN-003Success Metrics and Product EvidenceSuccess Metrics and Product Evidence: scope, decisions, requirements, evidence, risks, and traceability for the Agent OS programme.VisionIOS-VS-ANDROID-VS-AGENT-OSIos Vs Android Vs Agent OsResearchPRIOR-ART-ATLASPrior Art AtlasResearch
Execution
Current engineering work.
Active work items connect implementation, experiments, decisions and acceptance evidence to the specifications they advance.
#71 Waitlist hardening follow-ups from multi-model review (honeypot false-positive, Retry-After mismatch) open #70 site-validate CI: corpus-sync check fails nondeterministically on Linux (readdirSync order) open #68 Correct HyperOS repo references to AgentOS in portal metadata files closed #64 [AOS-MICROAPP-000] Text-to-Micro-App and cross-surface runtime open #63 [AOS-MESH-000] Agent Mesh and off-grid peer connectivity open